The Waitomo Caves attract approximately 500 000 tourists each year. A requirement of tourist cave management is that the partial pressure of carbon dioxide (PCO₂) is kept below levels that are: hazardous to the health of visitors, hazardous to the glowworms and other natural inhabitants, or potentially corrosive to speleothems. For the Glowworm Cave at Waitomo, the maximum permissible PCO₂ level is 2400 ppm. When exceeded, the tourist operators are required to close the cave. Ten years of monitoring data at the Glowworm Cave was analysed. Most of the variation in PCO₂ could be attributed to CO₂ respired by tourists, and the mixing of cave air with lower PCO₂ outside air. Occasionally, there were periods with high PCO₂ levels while the cave was closed to tourists. The main objective of this study was to investigate the potential role of the Waitomo Stream in contributing CO₂ to the Glowworm Cave atmosphere. Analysis of ten years of Glowworm Cave monitoring data showed that the 2400 ppm PCO₂ limit was, on average, exceeded five times each year, with a total of 48 events between 1998 and 2007. Of the PCO₂ limit exceedences, approximately 31% of events were largely driven by high tourist numbers; 27% of PCO₂ limit exceedences were mainly driven by increased discharge, rainfall, and/or a low temperature gradient between the cave and outside air, whilst 29% of the PCO₂ limit exceedences were due to a combination of tourists and increased discharge, rainfall, and/or a low temperature gradient. The remaining 13% of exceedences were unexplained by tourists or the factors investigated. It may be that the unexplained exceedences were due to the night time closure of the cave door, restricting air exchange. The PCO₂ of the Waitomo Stream was measured by equilibrating air with the streamwater within a closed loop. The air was passed continuously through an infrared gas analyser (IRGA). The streamwater PCO₂ typically ranged between 600 - 1200 ppm. Fluctuations in the PCO₂ of the Waitomo Stream coincided with PCO₂ fluctuations in the Glowworm Cave air, and under most conditions, the stream probably acted as a sink for cave air CO₂. However, following rainfall events, the stream PCO₂ increased, exceeding cave air PCO₂, thus acting as a source of CO₂ to the cave air. High stream PCO₂ often occurred at times when air flow through the cave was restricted, e.g. when the temperature gradient between the cave air and outside air was low, or stream levels were high, thus limiting air movement. The combination of high stream PCO₂ and a low temperature gradient increased the likelihood of high cave air PCO₂. Dripwater was measured to determine whether an increase in dripwater PCO₂ occurred in response to rainfall events. When rainfall events resulted in increased discharge, the dripwater PCO₂ sometimes increased (occasionally exceeding 5000 ppm), however the pattern was not consistent. The chemistry of the Waitomo and Okohua (Ruakuri) Streams was monitored with daily samples collected and analysed for major ions: HCO₃ -, Ca²⁺, Na⁺ and Mg²⁺, and δ¹³C stable isotope. The HCO₃ -, Ca²⁺, Na⁺ and Mg²⁺ concentrations in the streamwater decreased with increased discharge, presumably due to dilution. Increased discharge following rainfall events correlated with increasing PCO₂ in the Waitomo Stream, suggesting that soil atmosphere CO₂ dissolved in soil waters, and carried to the stream by saturated flow, was responsible for the streamwater PCO₂ increase. Ca in the stream showed both an increase and a decrease with respect to rainfall. Increased Ca in the stream occurred at times when the discharged waters were coming from the phreatic zone, and thus sufficient time had lapsed for CO₂ in the discharge waters to react with the limestone (carbonate dissolution reaction). Decreased Ca occurred when the infiltration and percolation of rainwater was rapid, and thus the streamwater was characterised by a higher PCO₂ and a lower Ca concentration, as insufficient time had lapsed for the discharge waters to equilibrate with the limestone. Increased negativity in the δ¹³C of the Waitomo and Ruakuri Streams coincided with increased discharge. During summer low flow, the δ¹³C of Waitomo Stream waters was -11.3‰, whereas during high stream discharge events, the δ¹³C dropped to -12 - -14‰. The δ¹³C of limestone is 0‰, the atmosphere is -7‰, and the soil atmosphere is reported to be about -24‰, thus the decrease in δ¹³C during high flow events supports the contention that soil atmosphere CO₂ is a likely source of the increased CO₂ in flood waters.