diff --git a/configure.ac b/configure.ac
index 1558a4d..167ca69 100644
--- a/configure.ac
+++ b/configure.ac
@@ -151,7 +151,7 @@ AC_ARG_WITH(libao,
 
 sound="yes"
 AC_ARG_WITH(sound,
-    [  --with-sound            select sound system ("oss", "sgi", "sun" or "libao") ],
+    [  --with-sound            select sound system ("oss", "sgi", "sun", "alsa" or "libao") ],
     [ 
     sound="$withval" 
     ])
@@ -177,9 +177,13 @@ elif test "$sound" = libao; then
     SOUNDOBJ="rdpsnd.o rdpsnd_libao.o"
     LDFLAGS="$LDFLAGS -lao"
     AC_DEFINE(WITH_RDPSND)
+elif test "$sound" = alsa; then
+    SOUNDOBJ="rdpsnd.o rdpsnd_alsa.o"
+    LDFLAGS="$LDFLAGS -lasound"
+    AC_DEFINE(WITH_RDPSND)
 else
    AC_MSG_WARN([sound support disabled])
-   AC_MSG_WARN([Currently supported systems are Open Sound System (oss), SGI AL (sgi), Sun/BSD (sun) and libao])
+   AC_MSG_WARN([Currently supported systems are Open Sound System (oss), SGI AL (sgi), Sun/BSD (sun), ALSA (alsa) and libao])
 fi
 AC_SUBST(SOUNDOBJ)
 
diff --git a/rdpsnd_alsa.c b/rdpsnd_alsa.c
new file mode 100644
index 0000000..a72cb2d
--- /dev/null
+++ b/rdpsnd_alsa.c
@@ -0,0 +1,332 @@
+/* -*- c-basic-offset: 8 -*-
+   rdesktop: A Remote Desktop Protocol client.
+   Sound Channel Process Functions - alsa-driver
+   Copyright (C) Matthew Chapman 2003
+   Copyright (C) GuoJunBo guojunbo@ict.ac.cn 2003
+   Copyright (C) Michael Gernoth mike@zerfleddert.de 2006
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program; if not, write to the Free Software
+   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+*/
+
+#include "rdesktop.h"
+#include <unistd.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <alsa/asoundlib.h>
+#include <sys/time.h>
+
+#define DEFAULTDEVICE	"default"
+#define MAX_QUEUE	10
+#define MAX_FRAMES	32
+
+int g_dsp_fd;
+BOOL g_dsp_busy = False;
+static snd_pcm_t *pcm_handle = NULL;
+static snd_pcm_stream_t stream = SND_PCM_STREAM_PLAYBACK;
+static BOOL reopened;
+static short samplewidth;
+static int audiochannels;
+
+static struct audio_packet
+{
+	struct stream s;
+	uint16 tick;
+	uint8 index;
+} packet_queue[MAX_QUEUE];
+static unsigned int queue_hi, queue_lo;
+
+BOOL
+wave_out_open(void)
+{
+	char *pcm_name;
+	int err;
+
+	pcm_name = xstrdup(DEFAULTDEVICE);
+
+	if ((err = snd_pcm_open(&pcm_handle, pcm_name, stream, 0)) < 0)
+	{
+		error("snd_pcm_open: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	g_dsp_fd = 0;
+	queue_lo = queue_hi = 0;
+
+	reopened = True;
+
+	return True;
+}
+
+void
+wave_out_close(void)
+{
+	/* Ack all remaining packets */
+	while (queue_lo != queue_hi)
+	{
+		rdpsnd_send_completion(packet_queue[queue_lo].tick, packet_queue[queue_lo].index);
+		free(packet_queue[queue_lo].s.data);
+		queue_lo = (queue_lo + 1) % MAX_QUEUE;
+	}
+
+	if (pcm_handle)
+	{
+		snd_pcm_drop(pcm_handle);
+		snd_pcm_close(pcm_handle);
+	}
+}
+
+BOOL
+wave_out_format_supported(WAVEFORMATEX * pwfx)
+{
+#if 0
+	int err;
+	snd_pcm_hw_params_t *hwparams = NULL;
+
+	if ((err = snd_pcm_hw_params_malloc(&hwparams)) < 0)
+	{
+		error("snd_pcm_hw_params_malloc: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	if ((err = snd_pcm_hw_params_any(pcm_handle, hwparams)) < 0)
+	{
+		error("snd_pcm_hw_params_malloc: %s\n", snd_strerror(err));
+		return False;
+	}
+	snd_pcm_hw_params_free(hwparams);
+#endif
+
+	if (pwfx->wFormatTag != WAVE_FORMAT_PCM)
+		return False;
+	if ((pwfx->nChannels != 1) && (pwfx->nChannels != 2))
+		return False;
+	if ((pwfx->wBitsPerSample != 8) && (pwfx->wBitsPerSample != 16))
+		return False;
+	if ((pwfx->nSamplesPerSec != 44100) && (pwfx->nSamplesPerSec != 22050))
+		return False;
+
+	return True;
+}
+
+BOOL
+wave_out_set_format(WAVEFORMATEX * pwfx)
+{
+	snd_pcm_hw_params_t *hwparams = NULL;
+	unsigned int rate, exact_rate;
+	int err;
+	unsigned int buffertime;
+
+	samplewidth = pwfx->wBitsPerSample / 8;
+
+	if ((err = snd_pcm_hw_params_malloc(&hwparams)) < 0)
+	{
+		error("snd_pcm_hw_params_malloc: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	if ((err = snd_pcm_hw_params_any(pcm_handle, hwparams)) < 0)
+	{
+		error("snd_pcm_hw_params_any: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	if ((err =
+	     snd_pcm_hw_params_set_access(pcm_handle, hwparams, SND_PCM_ACCESS_RW_INTERLEAVED)) < 0)
+	{
+		error("snd_pcm_hw_params_set_access: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	if (pwfx->wBitsPerSample == 16)
+	{
+		if ((err =
+		     snd_pcm_hw_params_set_format(pcm_handle, hwparams, SND_PCM_FORMAT_S16_LE)) < 0)
+		{
+			error("snd_pcm_hw_params_set_format: %s\n", snd_strerror(err));
+			return False;
+		}
+	}
+	else
+	{
+		if ((err =
+		     snd_pcm_hw_params_set_format(pcm_handle, hwparams, SND_PCM_FORMAT_S8)) < 0)
+		{
+			error("snd_pcm_hw_params_set_format: %s\n", snd_strerror(err));
+			return False;
+		}
+	}
+
+#if 0
+	if ((err = snd_pcm_hw_params_set_rate_resample(pcm_handle, hwparams, 1)) < 0)
+	{
+		error("snd_pcm_hw_params_set_rate_resample: %s\n", snd_strerror(err));
+		return False;
+	}
+#endif
+
+	exact_rate = rate = pwfx->nSamplesPerSec;
+	if ((err = snd_pcm_hw_params_set_rate_near(pcm_handle, hwparams, &exact_rate, 0)) < 0)
+	{
+		error("snd_pcm_hw_params_set_rate_near: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	audiochannels = pwfx->nChannels;
+	if ((err = snd_pcm_hw_params_set_channels(pcm_handle, hwparams, pwfx->nChannels)) < 0)
+	{
+		error("snd_pcm_hw_params_set_channels: %s\n", snd_strerror(err));
+		return False;
+	}
+
+
+	buffertime = 500000;	/* microseconds */
+	if ((err =
+	     snd_pcm_hw_params_set_buffer_time_near(pcm_handle, hwparams, &buffertime, 0)) < 0)
+	{
+		error("snd_pcm_hw_params_set_buffer_time_near: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	if ((err = snd_pcm_hw_params(pcm_handle, hwparams)) < 0)
+	{
+		error("snd_pcm_hw_params: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	snd_pcm_hw_params_free(hwparams);
+
+	if ((err = snd_pcm_prepare(pcm_handle)) < 0)
+	{
+		error("snd_pcm_prepare: %s\n", snd_strerror(err));
+		return False;
+	}
+
+	reopened = True;
+
+	return True;
+}
+
+void
+wave_out_volume(uint16 left, uint16 right)
+{
+	static int warned = 0;
+
+	if (!warned)
+	{
+		warning("volume changes currently not supported with experimental alsa-output\n");
+		warned = 1;
+	}
+}
+
+void
+wave_out_write(STREAM s, uint16 tick, uint8 index)
+{
+	struct audio_packet *packet = &packet_queue[queue_hi];
+	unsigned int next_hi = (queue_hi + 1) % MAX_QUEUE;
+
+	if (next_hi == queue_lo)
+	{
+		error("No space to queue audio packet\n");
+		return;
+	}
+
+	queue_hi = next_hi;
+
+	packet->s = *s;
+	packet->tick = tick;
+	packet->index = index;
+	packet->s.p += 4;
+
+	/* we steal the data buffer from s, give it a new one */
+	s->data = (uint8 *) malloc(s->size);
+
+	if (!g_dsp_busy)
+		wave_out_play();
+}
+
+void
+wave_out_play(void)
+{
+	struct audio_packet *packet;
+	STREAM out;
+	int len;
+	static long prev_s, prev_us;
+	unsigned int duration;
+	struct timeval tv;
+	int next_tick;
+
+	if (reopened)
+	{
+		reopened = False;
+		gettimeofday(&tv, NULL);
+		prev_s = tv.tv_sec;
+		prev_us = tv.tv_usec;
+	}
+
+	if (queue_lo == queue_hi)
+	{
+		g_dsp_busy = 0;
+		return;
+	}
+
+	packet = &packet_queue[queue_lo];
+	out = &packet->s;
+
+	if (((queue_lo + 1) % MAX_QUEUE) != queue_hi)
+	{
+		next_tick = packet_queue[(queue_lo + 1) % MAX_QUEUE].tick;
+	}
+	else
+	{
+		next_tick = (packet->tick + 65535) % 65536;
+	}
+
+	len = (out->end - out->p) / (samplewidth * audiochannels);
+	if ((len = snd_pcm_writei(pcm_handle, out->p, ((MAX_FRAMES < len) ? MAX_FRAMES : len))) < 0)
+	{
+		snd_pcm_prepare(pcm_handle);
+		len = 0;
+	}
+	out->p += (len * samplewidth * audiochannels);
+
+	gettimeofday(&tv, NULL);
+
+	duration = ((tv.tv_sec - prev_s) * 1000000 + (tv.tv_usec - prev_us)) / 1000;
+
+	if (packet->tick > next_tick)
+		next_tick += 65536;
+
+	if ((out->p == out->end) || duration > next_tick - packet->tick + 500)
+	{
+		prev_s = tv.tv_sec;
+		prev_us = tv.tv_usec;
+
+		if (abs((next_tick - packet->tick) - duration) > 20)
+		{
+			DEBUG(("duration: %d, calc: %d, ", duration, next_tick - packet->tick));
+			DEBUG(("last: %d, is: %d, should: %d\n", packet->tick,
+			       (packet->tick + duration) % 65536, next_tick % 65536));
+		}
+
+		/* Until all drivers are using the windows sound-ticks, we need to
+		   substract the 50 ticks added later by rdpsnd.c */
+		rdpsnd_send_completion(((packet->tick + duration) % 65536) - 50, packet->index);
+		free(out->data);
+		queue_lo = (queue_lo + 1) % MAX_QUEUE;
+	}
+
+	g_dsp_busy = 1;
+	return;
+}