Accurate abundance estimates of plant populations are fundamental to numerous ecological questions and for conservation. Estimating population parameters for rare or cryptic plant species, however, can be challenging and thus developing and testing new methods is useful. We assessed the efficacy of distance sampling for estimating abundance and habitat associations of the endangered Pima pineapple cactus (Coryphantha scheeri var. robustispina), a rare plant in the Sonoran Desert of southwestern North America that has traditionally been surveyed with census-based methods. Distance sampling (DS) involves measuring distances between focal objects and samples of lines or points, and modeling detection functions that adjust estimates for variation in detection probability (P). Although often used in animal systems, DS remains largely untested for plants. We encountered 105 live individuals along 36.9 km of transects in 11 study plots placed across much of the geographic range of the species, and estimated an average density of 1.47 individuals/ha (CV = 0.139). Compared to values from intensive censuses, density estimates from DS were underestimated by only 2.3% on average and highly correlated on the untransformed (r = 0.84) and logarithmic (r = 0.93) scales. Estimates of P averaged 0.49 and declined as soils became increasingly dominated by larger soil substrates, and somewhat with increasing vegetation volume and decreasing cactus height. Local densities increased with increasing slope and soil substrate size and decreased with increasing vegetation volume (p ≤ 0.024). Combined with careful survey design, DS offers an efficient method for estimating population parameters for uncommon and cryptic plants.