Segmentation is an important step in systematic roadway safety analysis. Several segmentation strategies have been used for higher functional class roadways (e.g. freeways, high volume corridors, etc.), but little work or reasoning is available to provide guidance for segmenting low-volume roads. This study investigated the effect of three segmentation strategies on safety analysis for two-lane rural primary roads and secondary low volume rural roads (LVRRs). The Iowa crash and road databases were used and a sensitivity analysis was conducted to test the effect of segmentation strategies on the identification of high crash locations. The effect of carrying fixed segment length on secondary LVRRs was also analyzed. Due to their relatively higher volume, results for rural primary roads indicate that half-mile segments produce candidate lists that are manageable in size for road agencies to identify statistically significant crash locations. Due to relatively lower volumes and crash frequencies, a similar recommendation cannot be made for an effective fixed segment length for secondary LVRRs. However, the analysis does indicate that segment length does significantly impact the identification of high crash locations for these roads. Care is advised when developing candidate lists for low volume roads based on segmented systems where systemic treatment may be more appropriate.