Previous studies suggested that variability indices based on information extracted from the acoustic signal are potentially useful in assessing dysarthric speech. Because of the ease of data collection, this method is especially applicable in the clinical setting. This study assessed the relationship between variability indices of sentence repetitions obtained by functional data analysis (FDA) with intelligibility ratings and maximum performance tasks in groups of speakers with hypokinetic dysarthria and ataxic dysarthria. The results showed significant correlations between selected parameters, which varied with dysarthria type. Whereas for the speakers with ataxic dysarthria, the variability measure mainly reflected differences in intelligibility, for the group with hypokinetic dysarthria, there was a stronger relationship between variability indices and diadochokinetic (DDK) performance. Lack of stronger correlations between variability measures and intelligibility ratings and maximum performance tasks are possibly attributable to heterogeneity of severity across and within speaker groups. This study provides further evidence that variability measures such as the FDA might be sensitive to speech performance of speakers with dysarthria and can potentially differentiate between dysarthria types.