The purpose of this study was to investigate the long-term effects of a Rhythmic Speech Cuing (RSC) protocol on severe speech impairments resulting from Parkinson's disease (PD). Five individuals diagnosed with idiopathic PD and speech deficiencies resulting from PD participated in the study. Each participant received 12 treatment sessions of RSC over a 4-week period. Each 25-minute treatment session consisted of metered rhythmic cuing with fading (for 20 minutes) and conversational speech (for 5 minutes). For rhythmic cuing, participants hand-tapped and spoke each word syllable to an auditory metronome cue that was set to 60% of their habitual speech rate (HSR). Data were collected during three testing sessions: pre-, mid-, and post-test. Each 20-minute testing session consisted of the participant speaking a total of 22 sentences (i.e., 220 words) from the Assessment of Intelligibility of Dysarthric Speech (AIDS) (Yorkston & Beukelman, 1981a) at their own preferred pace with no cuing. A picture description task (PDT) was also used, in which participants reported what they observed in a color photo depicting an everyday scenario. The dependent variables observed via the AIDS were percentage of intelligible speech, speech rate (including duration and words per minute, or WPM), and intelligible words per minute (IWPM). Unintelligible words per minute (UWPM) and communication efficiency ratio (CER) were also observed with the AIDS. Additionally, intelligible words and content of speech were observed via the PDT. The results showed that RSC had a statistically significant main effect on content of speech scores. Although scores for only one dependent variable achieved statistical significance, scores for six out of eight dependent variables generated large effect sizes, thus indicating practical significance. Findings revealed that the current RSC treatment protocol was most effective in the treatment of individuals with severe speech deficiencies. The participant whose speech was the most severely affected demonstrated marked improvements in all dependent variables in the current research. For example, his percentage of intelligible speech increased from 65% to 89%, which is a percentage that is comparable to typical speakers. Moreover, his CER increased from 0.14 to 0.28, showing his communication efficiency doubled from pre- to post-test. For this participant, spontaneous speech samples via the PDT indicated that scores for both intelligible words and content of speech increased from pre- to post-test. These increases suggest skill transfer may have occurred from training to spontaneous speech. Limitations within the current research included difficulty assessing speech deficiencies in the screening procedure, and a small and varied sample. Therefore, clinicians should be vigilant in utilizing an assessment tool that will accurately measure level of speech impairment. A precise tool will be helpful in identifying individuals with severe deficiencies who are most likely to benefit from use of this treatment protocol. The current data demonstrate that within a one month RSC protocol, patients with severe speech impairments resulting from PD can maintain or improve functional speech.