Identifying dairy cattle experiencing heat stress can improve welfare and profitability. However, it is unclear how often we need to observe the animals for this purpose, nor the relevance of specific aspects of the heat stress response, particularly in terms of panting (e.g., closed vs. open mouth panting). Our objectives were to refine and validate methodology used to assess heat load in dairies by (1) determining sampling intervals to measure usage of heat abatement resources (HAR; sprayed water or shade); and (2) evaluating the relationship between respiration rates (RR) and panting characteristics (PC; drooling, open mouth, and tongue outside the mouth). High-producing lactating cows (>40 kg milk/d) were chosen from 4 California dry-lot dairies (8 cows/dairy, n = 32) and observed for at least 6 h (1000 to 1800 h, excluding milking) when air temperature, humidity and the combined index averaged 33°C, 30% and 79, respectively. Use of HAR by individual cows was recorded continuously. Every 5 min, RR and presence/absence of PC (observed for 10 s at a time) were recorded. For use of HAR, estimates from different sampling intervals (1, 5, 10, 15, 20, 30, 60, 90 and 120 min) were derived from continuous data. Linear regression was used to compare these estimates against the continuous measurements. Sampling intervals were considered accurate if they met 3 criteria: R² ≥ 0.9, intercept = 0 and slope = 1. The relationship between RR and each PC was analyzed using mixed models with random effects for cow and cow × PC, to generate appropriate error degrees of freedom. When the sampling interval for HAR was ≤ 60 min, R² values were ≥ 0.9 and both the intercept and slope did not differ from 0 and 1, respectively (P > 0.05). RR was higher when PC were present (least squares means ± standard error, breaths/min: with vs. without drool present, 97 ± 3 vs. 74 ± 3; open vs. closed mouth, 100 ± 4 vs. 80 ± 3; tongue outside vs. inside the mouth, 96 ± 4 vs. 81 ± 3; P < 0.001). In summary, use of HAR is best measured when using sampling interval ≤ 60 min. Panting was accompanied by higher RR, and further analysis is planned to determine appropriate sampling intervals for these measures.