Expenditure and economic impact analyses are common in outdoor recreation and tourism, but there are potential sources of error in resulting estimates. The present study extends expenditure evaluation using national period-based reporting data from Sweden. This appears to be the first study that: (i) systematically provided respondents with opportunities to correct reported expenditure amounts and (ii) utilized corrections, certainty evaluations, and survey completion time to adjust expenditure estimates. Greater expenditure complexity was associated with lower respondent certainty in expenditure reports. Survey completion time was positively correlated with certainty, though only marginally so. Fewer than 10% of respondents changed expenditure reports, and the changes did not have a dramatic effect on mean expenditure. Expenditure amounts in the fuel and grocery categories were the most likely to be changed and the most likely to affect reported certainty; this is consistent with potential error arising from allocation of expenditure to recreation versus non-recreation purposes. Most of the data treatments resulted in means and standard deviations that were within 10% of the common Naïve approach; however, the "high certainty" treatment substantially reduced the expenditure mean and standard deviation. Given the prevalence and importance of economic impact estimates, the diversity of analysis contexts, and the paucity of past research, further evaluation of alternate survey administration and data treatment approaches is recommended. This research will contribute to more informed policy making and management through enhanced data quality. Management implications Recreation is one of the important ecosystem services that natural areas provide, and the economic impact of visitor expenditure is one rationale for promoting outdoor recreation participation and sustaining natural areas. This study contributes to informed policy making and management by (i) providing guidance regarding data collection methods, including the role of probing and survey completion time, and (ii) illustrating different treatments of reported expenditure, with potential implications for data quality and results.