Just wanted to make an update to this thread. As was shown in the studies above, chloroform was not found to be mutagenic and it was also found to be removed from the body after inhalation fairly quickly. Most studies (see this one, this one, this one and this one) show that the linear dose model is not appropriate for predicting cancer risk for chloroform and that the weight of evidence indicates that chloroform is acting through a nongenotoxic-cytotoxic mode of action. That is, chloroform doesn't cause cancer through probabilistic mutations but rather through toxic effects at high levels. This means that below some cutoff, both toxic and cancer effects cease. In the last linked study, the 95% confidence level for lifetime excess cancer risk for continuously exposed human populations is 2200 ppb in air and 13,100 ppb in water. In practical terms, this would make chloroform a non-issue in both drinking water and pool/spa water.
As noted in this EPA link and in this article, the EPA changed their risk assessment for chloroform though have yet to update their inhalation risk estimates [EDIT] (further details are in this document) [END-EDIT]. The EPA Drinking Water Contaminants Limits are currently 60 ppb for haloacetic acids (HAA5) and 80 ppb for Total Trihalomethanes (TTHMs) with separate goals (not legal limits) of 0 for bromodichloromethane, 0 for bromoform, 60 ppb for dibromochloromethane, 70 ppb for chloroform, 0 for dichloroacetic acid, 20 ppb for trichloroacetic acid, 70 ppb for monochloroacetic acid and no goals for bromoacetic acid and dibromoacetic acid. In light of the science, the limit on chloroform may be excessively aggressive (note that the German DIN 19643 is even more aggressive with a target of 20 µg/L (ppb) for TTHMs). [EDIT] This document is a bit older but summarizes a large number of tests that imply a No Observed Adverse Effect Level (NOAEL) on the order of 10 ppm (10,000 ppb) or 1 mg/kg/day which with 100 ppb chloroform in water implies a 50 kg person drinking (or absorbing) 500 liters of water per day to reach the NOAEL (and the observed effects limit is higher than that). [END-EDIT]
The reason I bring this up is that I have been doing water sample studies of TTHMs in my pool and finding that while the brominated THMs were relatively low within a factor of 2 of tap water (which was 8-9 µg/L), the chloroform readings have been an order of magnitude higher than tap water (which was 18-23 µg/L). Since chloroform is volatile and the pool is usually covered, I'm going to see if having the pool open and aerating for one week during the winter reduces the chloroform concentration. Some water treatment facilities use aeration to reduce volatile disinfection by-products. I suspect that in outdoor pools that are uncovered more frequently that the volatiles are reduced, not only by exposure to the air but to sunlight as well (my automatic electric safety cover is mostly opaque). This data may also indicate that using chloroform or TTHMs as a proxy for the more serious mutagenic brominated THMs may be a flawed approach and overly aggressive. Specific testing of brominated THMs or, at a minimum, bromide levels would probably be more appropriate. It also raises concern for the use of bromine, especially relative to using chlorine. The Barcelona water had more bromide than most, but I would expect explicitly using bromine as the sanitizer would produce far higher brominated THM levels. This study of Barcelona chlorine and bromine pools shows that the brominated pools had far higher levels of bromoform though lower levels of chlorinated THMs. Even so, the predicted cancer risk of 75 µg/L bromoform if in drinking water at 2 liters per day for a lifetime would be a 1 in 50,000 increased cancer risk (EPA limits tend to be set at a 1 in 1,000,000 increased cancer risk -- also, these risk assessments include significant margin of exposure factors of at least 100).