During variable distance point counts a stationary observer records the species and distance for all birds detected during a fixed time interval. Such counts, repeated throughout the area of interest, can be used to estimate bird density so long as the following assumptions are made: (1) the distribution of birds is unaffected by the observer; (2) observers are certain of detecting near birds; (3) there is no error in measurement or estimation of distances; and (4) birds are stationary. Data collected during the present study indicated that at least two of these assumptions are not satisfied: estimates of distance based on sound are inaccurate and observers do not always detect near birds. No effect of observer presence on bird distribution was detected nor did there appear to be any significant movement of birds during counts. We recommend that bird counts be based only on sight detections and that, when bird movement is significant, instantaneous counts be used.