The present study sought to confirm melodic expectancy patterns discovered in a previous investigation and to determine whether data would be affected by altering the rhythmic condition of the stimuli. The three problems of this investigation were to study expectancies generated by two-note stimulus intervals of equal duration; the expectancies generated by stimulus intervals presented with a long-short rhythm; and the expectancies generated by stimulus intervals presented with a short-long rhythm. Fifty subjects were asked to sing what they believed would be the natural continuation of the melody begun by a two-note stimulus interval. The stimulus intervals were grouped in rhythmic sets, one set of neutral-rhythm stimuli, one set of long-short stimuli, and one set of short-long stimuli. The interval from the second note of the stimulus interval to the first sung note was transcribed as the generated expectancy response interval. The data were examined in two basic ways, using response data as a whole and examining data for each stimulus interval separately. A third method of data evaluation concerned harmonic triads occurring in the response data. Both when considering response frequency as a whole, and when considering response data separately for each melodic beginning, no significant difference associated with rhythmic condition could be found. Smaller response intervals were generated much more often than large intervals. Some stimulus intervals, notably the major second ascending, were observed to be much more powerful generators than others. It was concluded from these results that in response to two-note stimulus intervals melodic expectancy can clearly be shown to operate, confirming the results of an earlier study, but that no effect of rhythm on melodic expectancy can be shown to operate.