Objective: The aim of this study was to analyze the temporal evolution in the intervals between consultation, diagnosis, and treatment of leukemias and lymphomas in children and adolescents in São Paulo State, Brazil, from 2000 to 2023, identifying patterns of periodicity, temporal trends, and possible seasonality. Methods: An ecological time series study was conducted using data from 77 institutions registered with the Oncocenter Foundation of the State of São Paulo, grouped according to the International Classification of Childhood Cancer. Monthly median times between consultation, diagnosis, and treatment were calculated. Kwiatkowski–Phillips–Schmidt–Shin tests were applied to assess trends and stationarity; Seasonal Effects and Autocorrelation Structure was applied for the assessment of seasonality; and the Ljung–Box test was employed for autocorrelation, supporting the application of Autoregressive Integrated Moving Average models. Results: The time series analysis showed significant trends in the median times between consultation and diagnosis and between diagnosis and treatment for both groups of neoplasms analyzed (p < 0.01). Significant autocorrelation was observed for leukemia and myeloproliferative diseases, as well as lymphomas and reticuloendothelial neoplasms, with no indication of seasonality. Autoregressive models indicated robust temporal patterns necessary to predict the series' behavior over the study period (2000–2023). Conclusion: This study identified significant trends in the times between consultation, diagnosis, and treatment for childhood leukemia and lymphomas, with no evidence of seasonality. These findings highlight the need for faster interventions in pediatric oncology care and reinforce the importance of robust databases to guide public policies aimed at early detection and appropriate treatment.