Excluding all records in a many-to-many relationship

[ultimateguy]

I have a trivia database and I want to keep track of what questions were used at different establishments, so that I can avoid using the same question more than once at a single establishment.

Say for example, Question 1 was used at both pub A and pub B. I then want to build a query that shows me all of the questions that have NOT been used at pub B. I build the query seen above and in the Criteria for the ClientName field I put 'not "pub B" '. The problem with this query is that Question 1 has also been used at pub A, so it will return Question 1 in the results - not what I want given that Question 1 has already been used at pub B.

Basically I would like to know how to build a query that excludes ALL of the question ID's that have any association with pub B. I have tried building a query with a subquery, but gotten some wonky results.

[June7]

Try:

Query1
SELECT CompletedTrivias.CustomerID, UsedQuestions.QuestionID
FROM CompletedTrivias RIGHT JOIN UsedQuestions ON CompletedTrivias.ID = UsedQuestions.TriviaID;

Query2
SELECT Clients.ID, Questions.ID
FROM Clients, Questions;

Query3
SELECT Query2.Clients.ID, Query2.Questions.ID, Query1.QuestionID
FROM Query2 LEFT JOIN Query1 ON (Query2.Questions.ID = Query1.QuestionID) AND (Query2.Clients.ID = Query1.CustomerID)
WHERE (((Query2.Clients.ID)=[enter client ID]) AND ((Query1.QuestionID) Is Null));

[ultimateguy]

Thank you, that works perfectly. I will need to work on becoming more adept at using joins.

[June7]

And when not to use joins. Note that Query2 does not have a join clause (these two tables do not a natural relationship and couldn't be joined anyway). This results in a Cartesian relationship where every record in each table joins with every record of other table to provide a dataset of all possible Client/Question pairs. Then Query3 is essentially a FindUnmatched to determine which of the pairs does not occur in Query1.