Based on the monthly precipitation, monthly mean air temperature and monthly mean runoff data of the Parana River Basin from 1951-2014, linear regression and Mann-Kendall method are adopted to analyze the tendency of time series, meanwhile, wavelet analysis is used to analyze the multi time scale characteristics of time series and the inner link between meteorological factors and hydrological factors. The result suggested that during the study period, the three factors of precipitation, temperature and runoff all showed a rising trend to different degrees, the temperature and runoff increased significantly, whereas the precipitation increased insignificantly. On the monthly scale, the three factors all had a main cycle of about 12 months, however they all had different sub cycles. Among them, air temperature had a weak sub cycle of about 6 months, runoff had sub cycles of about 36 months, 48 months, 64 months and 100 months, whereas monthly mean precipitation had no significant sub cycle. On the annual scale, there was no significant main cycle for all three factors, but precipitation had three sub cycles of about 3 a, 4 a, 8 a and 10 a, runoff had the same sub cycles as the precipitation, temperature had no significant and continuous sub cycle. Wavelet coherence spectrum results showed that monthly precipitation had a greater impact on the runoff of the Parana River Basin than monthly mean air temperature. The approach combining trend analysis with wavelet analysis can be applied to the analysis of the change patterns of hydro meteorological variables and their inner links on basin scale with satisfactory results.