Precise costing of vegetation surveys has become a necessity for parastatal research organizations. Competitive and cost‐effective budgets have to be tabled by these parastatals to secure funding for mapping and classification projects. The mean time required to record floristic data in the Sourish Mixed Bushveld, Transvaal, South Africa, using a recently developed, point‐canopy intercept method (PCIM) and a quadrat method (QM) was 20.5 minutes per stand (SD=5.8) and 45.6 minutes per stand (SD ‐ 11.9) respectively for the whole survey (n = 75 stands). Less time per species encountered was spent using the QM than the PCIM, but because the whole stand (circle with radius of 8 m) was sampled using the QM, more species per stand were recorded for the QM (mean = 60.4; SD ‐ 14.3) than for the PCIM (mean = 16.4; SD = 4.7). Linear regressions of species number versus time taken for the beginning and middle and end of the survey data sets were generated for both methods with correlation coefficients (r) of 0.900 and 0.901 respectively. These models will provide a basis for managers to budget more reliably.