In this study we compared the photosensitizer concentration in two experimental murine tumors using an in situ fluorescence detection instrument to examine temporal and spatial variations, after intravenous versus intratumor injection. Also, the variations in the estimate as detected by large area sampling and micro-region sampling are compared, in order to determine what the inter-tissue and inter-animal variations are, and how the method of sampling affects this estimate. The latter study was carried out ex vivo in the same tumors, which had been harvested and frozen after in vivo measurements were made. The photosensitizer, disulphonated aluminum phthalocyanine (AlPcS2) was injected either intravenously (IV) or directly into the tumor (ITu), using two murine models, MTG-B (mammary adenocarcinoma) and RIF-1 (radiation-induced fibrosarcoma) grown subcutaneously on the flank. An in situ microsampling fluorescence probe was used to assess photosensitizer concentration, through real-time measurement of the remitted intensity. The photosensitizer concentration was evaluated at 8 time endpoints between 15 min and 48 h post-injection. Inter-tumor and intra-tumor variations were assessed by repeated samples from the tumor tissues. The average photosensitizer level reaches a peak between 3 to 6 h in both tumor and normal tissues using IV administration, but peaks within 1 h following ITu administration. MTG-B tumors demonstrated a factor of 2 higher uptake than RIF-1 tumors. The pharmacokinetic uptake rates of the RIF-1 tumor were 3 times faster than for MTG-B, while there was no statistical difference in their clearance rates. Preferential uptake of AlPcS2 by both tumors compared to contra-lateral flank subcutaneous normal tissue was documented, with ITu injection exceeding IV injection by a factor of 10 in the tumor to normal tissue ratio. Inter-animal standard deviation in the mean fluorescence was near 76% for both routes of administration, but estimates of the variation within tumor were near 16% standard deviation when a large sampling volume was used. In contrast, microscopic intra-tumor standard deviation in the mean estimate was near 76%, with IV injection, indicating that high heterogeneity exists in the photosensitizer concentration on a smaller distance scale. The inter-tumor variation was reduced by ITu injection, but at the expense of increasing intra-tumor variation.